Unverified60% confidenceFactExact time
The Jetson Orin Nano (8GB version) consumes only 15W of power and can run quantized 2B models smoothly.
2
Sources
60%
Confidence
Long-term
Relevance
9/9/2026
First Seen
Sources
Related Entities
Related Claims
VerifiedThe NVIDIA Jetson Nano has a compute performance of 472 GFLOPS and a power consumption of 5–10 W; the Jetson AGX Orin has a compute performance of 275 TOPS and a power consumption of 15–60 W.73% similarUnverifiedThe Jetson Orin series offers up to 64 GB of unified memory paired with LPDDR5 high bandwidth.67% similarUnverifiedThe Jetson Nano running lightweight models such as YOLOv5s or YOLOv8n typically achieves 10–25 FPS after TensorRT optimization66% similarUnverifiedOn a 16 GB M4 Mac Mini, after loading Ornith 9B GGUF, 6.2 GB is used for model weights, and the maximum context length can reach 260,000 tokens.65% similarUnverifiedJetson Xavier NX的功耗仅为10-15W64% similar
Cite This Claim
Stable URI
https://kongchang.com/claim/882140API
curl https://kongchang.com/api/v1/knowledge/claims/882140MCP
get_claim(id=882140)